ABSTRACT The overarching goal of this research is to promote children's positive, adaptive pathways of socioemotional development and to prevent maladaptive pathways. Our work explains why some children embark on positive paths toward prosocial, internalized, rule-abiding conduct, and robust social competence, whereas others enter maladaptive paths toward callousness, disregard for conduct rules and others' feelings, antisocial behavior, and impoverished competence. We focus on the parent-child early attachment relationship, formed in the first years of life, as a source of those divergent pathways, and we longitudinally chart its complex, indirect yet powerful, long-term implications. On the basis of our past and current work, correlational and experimental, in low- and high-risk families, we propose that although early parent-child attachment may not have long-term unqualified, direct effects, it nevertheless serves as a powerful catalyst or moderator of future dynamics unfolding between the parent and the child. Specifically, early attachment insecurity sets the stage for an adversarial, negative cascade. In insecure parent-child dyads, the child's difficult temperament easily triggers the parent's negative, coercive, power-assertive control, which, in turn, leads to maladaptive child outcomes. In contrast, early security sets the stage for positive, cooperative, effective socialization, and defuses risks of negative cascades. To elucidate mechanisms explaining those processes, we propose that the divergent cascades are due to parents' and children's differing internal representations, expectations, and perceptions of each other (Internal Working Models, or IWMs) that characterize insecure and secure dyads. Those IWMs then come to guide parents' and children's behavior and interactions. A new study of 200 community mothers, fathers, and infants, intensively assessed at 7-9, 15-17, 36-38, and 46-48 months, will test this model. Deploying state-of-the-science measures of parents' and children's social representations, in Aim 1 we examine their emerging negative or positive IWMs of each other as linked to their early insecure or secure attachment relationships. Those IWMs are then tested as key mechanisms accounting for divergent cascades that unfold in insecure and secure relationships. In Aim 2, we examine the parent's IWM of the child as moderating the link between child difficulty and parental control, and in Aim 3, we examine the child's IWM of the parent as moderating the link between parental control and child outcomes. Our team includes experts in socioemotional development, infant and adult cognition, molecular genetics, and methodology and statistics. In a multi-method, multi-level approach, we collect observational, genetic, and reported measures of parent and child social cognition, temperament, attachment, parental control, and children's outcomes. Analyses rely on structural equation modeling to elucidate mechanisms of divergent developmental cascades unfolding over time. This research will produce a novel, long-advocated, but yet to be realized, synthesis of attachment, temperament, internal representation, and behavior in pathways to children's adjustment.